My #LearnML Story — An Attempt To Inspire Young ML Aspirers
#LearnML Because I learned machine learning, I now create applications that have a potential for large scale impact. I am Utkarsh, and I am a machine learning obsessed entrepreneur who wants to change the world.
Just two semesters into learning computer science at Georgia Tech, I came across Andrew Ng’s coursera course on machine learning. I was hooked. I binge studied it like watching friends, finishing it in a week.
In 2013, I implemented a modified version of the tf-idf algorithm to auto-grade email exchanges for the NGO IMentor, in New York. The algorithm, along with my team’s work in building an interface for it, won the J.P. Morgan Code for Good NY hackathon, and the system is currently in use by the NGO. The system has allowed the NGO to take on an exponentially larger number of children, scaling their impact tremendously.
Earlier this year, along with four other undergrads, I began work on a project for intractable epilepsy, in collaboration with a neurologist from Emory Hospital. The existing programs are highly inaccurate, leading to 8,000 faulty brain surgeries every year in the US alone i.e. 8000 children undergo brain surgery with no positive impact on their lives. We are attempting to make spike detection in EEG data autonomous and accurate, using wavelet transform and unsupervised clustering.
In October this year, I had been working on a premature baby incubator project at CERN. A doctor from the Geneva General Hospital popped in and started talking about temperature sensors, and measurement data. We spent the next 6 hours discussing the nature of the data and observed trends. I then spent the rest of the night developing several data models, presenting two to the doctor, who spent another few hours with me, fine-tuning parameters based on the data and based on his extensive experience in the field. I am now working in collaboration with CERN and ThePort organization to integrate the algorithm into a physical incubator crafted by my team, working with the government of Cameroon where the problem is most rampant. We hope to have a prototype deployed in the next 18 months, with the ultimate aim of saving 100,000+ lives a year.
My story is that of a young student, who came across machine learning just as he was delving into computer science. Machine learning is an obsession with me, simply because of the unlimited applications and the boundless potential of the subject. I love learning more and more about machine learning, and applying it to issues that really matter; to solutions that will have impact. I am grateful to Andrew Ng for putting the course out there for free, and for helping me realize the belief, that a computer scientist can in fact, change the world.
[Edit]: This story was written as part of an online competition to inspire people to learn Machine Learning. Here is the contest — http://blog.coursera.org/post/133537959852/join-coursera-co-founder-andrew-ng-in-inspiring, and here are the winning entries — http://blog.coursera.org/post/135215105522/andrew-ng-announces-winners-of-learnml-contest.